gjc-sdk-operate
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for the agent to interact with the
gjcCLI. It includes defensive programming requirements, such as mandating that all data be passed asargvvalues rather than via shell command strings, which prevents shell injection vulnerabilities. - [INDIRECT_PROMPT_INJECTION]: The skill acknowledges and manages the risks associated with processing external tool outputs.
- Ingestion points: Data enters the agent's context through the output of
gjcCLI commands (e.g.,session list,raw query). - Boundary markers: A mandatory, single-use human approval workflow is required before any command is executed, acting as a critical safety boundary.
- Capability inventory: The skill facilitates specific session management and control operations via the
gjcCLI, as documented in the allowlist. - Sanitization: The instructions require the use of structured JSON for machine communication and mandate SHA256 integrity checks when retrieving error evidence fragments.
Audit Metadata